Volumetric data is an increasingly important part of medical systems, be it visualization of computer tomography data, or the tissue representation of realistic and even patient specific surgery simulation. Most algorithms for haptic interaction with actively deformable tissues and data to date use more or less simple surface representations of the data. This does not suffice in many cases.
The aim of this project is to expand the knowledge of how tissues responds to haptic interaction and produce new data structures, algorithms and methods for effective implementation and approximation of data deformation. The project uses volumetric representations to allow for palpation and free exploration of the full extent of the 3D data.
Todays medical scanners are capable of producing increasily high resolution and high quality data of patients in less time than it takes to go for lunch. This data contains information about the patient that may be valuable in more ways than the conventional examination — searching for abnormalities on slice by slice.
Volume visualization is capable of showing these data in a manner very similar to how the tissues might look in reality. Some systems can even be adjusted to show a photorealistic rendering that is in some senses quite indistinguishable from opening the body of the real patient. An obvious application of this is of course surgery simulations on authentic patient data, for example for pre-surgical planning and training, general surgery training, or syndrome specific surgery training.
For this new methods are needed that allow for realistic deformations in the volumetric data. New methods for haptic interaction with such dynamically deformable volumetric data will also then be further used in minimally invasive blind operations, such as biopsy and needle insertion. The methods can also be applied in multi-modal data exploration of dynamic behaviours in both medical and scientific visualization.
This project developes algorithms for interaction with deformable and dynamic volumetric data. This can be used to simulate, for example, biopsy on authentic and patient specific data. We will also search for effective data structures and algorithms that allow for deformation and palpation of medical data. It is here important that the underlying tissue distribution is deflected through the haptic feedback on the surface, for example the skin.
The project will extend the data structures to allow for topological changes and so allow for cutting and thus simulation of surgery procedures on the patient specific data.
The continuous results from this project are available on the publications list and sometimes eventually also integrated to the Volume Haptics Toolkit.
Please contact Karljohan Lundin for more information.